Soot emissions from diesel internal combustion engines are strictly regulated nowadays. Laser extinction measurement (LEM) and natural luminosity (NL) of sooty flames are commonly applied to study soot. LEM measures soot along the laser beam path and it can probe soot regardless of temperature. NL integrates the whole field of view and relies on soot temperature. In this work, a comparison of simultaneously recorded LEM and NL data has been performed in a heavy-duty optical engine. A 685 nm laser beam is used for LEM. The laser was modulated at 63 kHz, which facilitated subtraction of the background NL signal from the raw LEM data. By Beer-Lambert’s law, KL factor can be calculated and used as a metric to describe soot measurements. A compensation of transmitted laser intensity fluctuation and soot deposits on optical windows has been performed in this work. The data compensation successfully reduced the transmitted laser intensity fluctuation and made it possible to study in-cylinder low temperature soot residual. The KL curves were compared with NL curve in this work. In the late cycle the KL curve can successfully show the low temperature soot which is not detected by NL. The KL curve is found to rise about 2 CAD ahead of the corresponding NL curve due to liquid fuel spray disturbance. In this case, LEM is not a suitable method to calculate KL for analyzing the early soot formation if there are liquid phase fuel droplets crossing the probing laser beam.